常见肺磨玻璃密度影的影像学特征分析及鉴别诊断

Analysis of imaging characteristics and differential diagnosis of common ground glass density shadow(GGO)

目的:分析肺磨玻璃密度影(Ground-glass Opacity,GGO)的影像学特点,对比CT传统读片与人工智能(AI)技术对GGO良恶性判断的诊断价值,掌握肺内常见GGO影像学特征及鉴别诊断的方法。方法:回顾性分析CT检查中发现的肺内GGO患者182例,良性GGO为126例,恶性为52例;对肺内GGO影像特点进行分析,以病理学和综合诊断为金标准,比较CT传统读片和人工智能(AI)技术判断肺磨玻璃密度影良恶性的诊断效能。结果:CT传统读片诊断肺内GGO灵敏度为93.7%,特异度为96.2%,诊断符合率为94.4%。人工智能(AI)技术诊断肺内GGO灵敏度为95.2%,特异度为88.5%,诊断符合率为93.3%。结论:CT传统读片结合人工智能(AI)技术综合诊断肺内GGO病变,有助于进行良恶性的鉴别诊断,能够提高诊断符合率,减少漏诊率,为临床早期诊疗提供方向。

Objective:Analyze the imaging characteristics of ground glass opacity(GGO),compare the diagnostic value of traditional CT film reading and artificial intelligence(AI)technology in judging the benign and malignant of GGO,and master the imaging characteristics and differential diagnosis methods of common GGO in the lung.Methods:182 patients with pulmonary GGO found in CT examination were analyzed retrospectively,including 126 cases of benign GGO and 52 cases of malignant GGO;The imaging characteristics of GGO in the lung were analyzed.Taking pathology and comprehensive diagnosis as the gold standard,the diagnostic efficacy of traditional CT reading and artificial intelligence(AI)technology in judging the benign and malignant of pulmonary ground glass density shadow was compared.Results:The sensitivity,specificity and coincidence rate of traditional CT reading in the diagnosis of pulmonary GGO were 93.7%,96.2%and 94.4%.The sensitivity and specificity of artificial intelligence(AI)in the diagnosis of pulmonary GGO were 95.2%,88.5%and 93.3%.Conclusion:Traditional CT film reading combined with artificial intelligence(AI)technology can comprehensively diagnose pulmonary GGO lesions,which is helpful for the differential diagnosis of benign and malignant,improve the diagnostic coincidence rate,reduce the missed diagnosis rate,and provide direction for early clinical diagnosis and treatment.